The design of experiments (DOE) is the study of how to structure an information-gathering exercise where variation is present. In statistics, controlled experiments are usually implied, meaning that study units are randomly assigned to conditions where levels of a variable have been independently ...
0
votes
0answers
11 views
What statistical test would I use to compare the effect of pre-existing attitudes on post intervention scores?
Basic design:
Asking people to indicate their attitude toward getting a flu shot (4 options: very negative, mostly negative, mostly positive, very positive)
Then half of the people will receive ...
0
votes
0answers
13 views
Test to check reliability of sample splitting procedure
As someone with little background in statistics, I am trying to test the reliability of the splitting procedure for screen analysis currently done in the lab (if the particle size distribution will be ...
0
votes
0answers
17 views
What experimental design should I use?
I am trying to investigate the impact of a ban on plastic bags on the impact of i) pollution levels and ii) the purchase of alternative sources of plastic bags (i.e. 'bags for life').
What would be ...
0
votes
0answers
9 views
post-hoc, pairwise PERMANOVA for factor with three levels
Goodmorning,
I have a mixed, multifactorial design with 4 factors. I've run my PERMANOVA which returned a significant result for fixed factor: year (3 levels). My thesis pertains to changes over time,...
0
votes
0answers
18 views
What is overlap of covariates? Why is it less of a concern in randomized experiments? [on hold]
I am actually learning experimental design. I'm a bit confused about the overlap covariates. Could someone help me on this problem?
6
votes
3answers
193 views
Regression and causality in econometrics
In regression in general and in linear regression in particular causal interpretation about parameters is sometimes permitted. At least in econometrics literature, but not only, when causal ...
0
votes
0answers
12 views
Correct group of factorial and repeated-measure ANOVA
I'm currently having trouble defining what type of ANOVA I need to run as well as how to run it correctly. This is the following setup:
I have 4 groups of 30 people doing a 2-session task (one after ...
0
votes
1answer
23 views
Algorithmic or structural limitations of space-filling Latin hypercube sampling
I'm new to Latin hypercube sampling, and am trying to understand if the somewhat odd sampling that results from the Matlab function lhsdesign is a limitation of the ...
1
vote
0answers
18 views
Why can you not bypass the strong ignorability/unconfoundness assumption via iterated expectations?
Suppose we have that $\left(Y(1), Y(0)\right)$ are potential outcomes with $X$ being the covariate and $Z$ the treatment assignment. Typically in causal inference, one will assume strong ignorability ...
1
vote
1answer
96 views
non stochastic regressors
In the multiple linear regression analysis if regressors are non-stochastic the causal interpretation of parameters is automatically permitted?
I think so, because it seems me that the model can be ...
0
votes
0answers
7 views
Parameter setting in minitab Question
I want to optimize parameters of a genetic algorithm in Minitab. I want to use Taguchi method. To do this, I created three random datasets (with similar seeds) and run each one five times. and ...
0
votes
0answers
23 views
When should we include the “Replicate” in an ANOVA table for Factorial Design?
When I have been learning how to do Factorial Designs ($2^k$ mainly) the ANOVA table would usually include the main effects, all the interaction effects if necessary, and the error. However, on one of ...
1
vote
0answers
40 views
Is this a valid experimental design?
An experiment to test 2 new treatments was performed in the following setup:
Animals were divided into 2 groups: young (A) and old (B). Animals were present for 3 periods. Each group consisted of 81 ...
1
vote
1answer
28 views
How to use Design of Experiment (DoE) to reduce the number of simulations?
I am planning to do simulation for parametric study and there are 9 parameters in total. I was suggested to use DoE to reduce the number of simulations that I need to do. I studied the basic of DoE ...
0
votes
0answers
7 views
Randomised controlled block experiment - choosing the blocks
I'm confused about how to decide blocking, and what that means for experimenting with it. I've seen other questions on the site and don't feel that they answer this, I think this is more basic.
For ...
0
votes
1answer
29 views
Experimental Analysis with Several Discrete Treatments
I am analyzing data that are originating from an randomized experiment and I am new to this. There was one control groups and three different treatment groups. The treatment groups are discrete and ...
0
votes
1answer
15 views
Difference between Multistage Sampling and Stratified Random Sampling?
I know the question is a very elementary one, but I simply cannot understand the difference other than the fact that an SRS is a form of Multi-Stage Sampling. Can anyone provide a simple example(s) to ...
0
votes
1answer
48 views
non stochastic regressors and causation
Randomized controlled experiment is base case for causality (also) in regression.
However currently I’m analyzing the role of causality in linear regression as shown in many econometrics textbook. ...
3
votes
1answer
62 views
Proof that random assignment guarantees the same probability for the individual to be selected?
Consider a completely randomized design, where every unit is randomly assigned to a treatment group. Let's say we have 30 observations and 3 treatment groups. When we choose one observation at random, ...
2
votes
1answer
37 views
When should a randomised Latin Square be used rather than a standard Latin Square
I'm looking at Latin Squares. I've seen standard Latin squares, and Latin squares when the rows are randomised, then columns are randomised.
When would one be used over the other?
It seems as ...
2
votes
0answers
29 views
How does the “BiasAdjust” term in Match() function from Matching R package work?
In the Match() function in the Matching R package, it takes in inputs like:
...
1
vote
1answer
20 views
For the Match() function in the R package, “Matching”, what algorithm is used for propensity score matching?
In the Matching package in R, one can conduct propensity score matching if propensity scores are passed to the ...
2
votes
2answers
26 views
In observational studies with propensity score matching, why is the ATT usually reported? Is the ATE usually not available?
In most propensity score literature on observational studies, the ATT is usually reported and sought after. However, the ATE is usually not reported and in some cases I've read papers that claim you ...
1
vote
1answer
29 views
What is the difference between cross-over and repeated-measures designs?
I know that a cross-over experiment (of any number of treatments/levels) necessarily has a repeated-measures design, but that not all repeated-measures designs are cross-overs. However, I wasn't able ...
1
vote
1answer
39 views
How to choose Control groups for Causal Impact algorithm?
I'm running an experiment and want to use the Causal Impact function to assess how well it performs.
I have 10 different cities. I'm looking to find out what is the best method for choosing which ...
0
votes
0answers
5 views
Collect validation evidence for an assessment tool
I want to collect evidence of validation for an assessment tool I have been working on. The tool is developed as a support tool for an expert evaluator to reduce the inherent subjective impact. I am ...
4
votes
2answers
89 views
In a randomized trial, what is the propensity score?
In Rosenbaum's 1983 paper, he states that "in a randomized trial, the propensity score is a known function so that there exists one accepted specification."
I am wondering what this specification is ...
1
vote
3answers
94 views
In observational studies, how can unconfoundedness, $(Y(1), Y(0)) \perp T \mid X$, hold if $(Y(1), Y(0))$ are fixed and non-random?
In observational studies, one can use the Rubin Causal Model to retrieve unbiased estimates, which usually there is a statement that is usually required which states that:
$$
(Y(1), Y(0)) \perp T \...
1
vote
2answers
48 views
What statistical tests can I use? Repeated measures design, two groups, each is control and experimental at different points in time
I need to determine if an intervention had an effect. In the experimental design, measures were taken at three points in time: pre-test, post-test #1 and post-test #2. One group (group A) received the ...
1
vote
1answer
18 views
Adding center points in $2^k$ models
I'm not sure I understand this concept. In $2^k$ designs, the independent variables have only two values and are coded as either being -1 (low value) or +1 (high value).
We can add a center point to ...
0
votes
1answer
27 views
Optimal Sample Size/Power Analysis for Panel Data
I'm running a economic experiment. From a previous pilot I already know which size of treatment-effect I can expect. Each participant plays 20 rounds, so it't a repeated measures approach. I want to ...
0
votes
0answers
20 views
Control group in a quasi natural experiment
I am doing a quasi-natural experiment in which I am looking at the implication of government policy in India on the performance of companies. Unfortunately, almost all of the companies in my sample ...
0
votes
0answers
26 views
In an observational study in causal inference, if $Y_i$, $T_i$, $X_i$ are outcomes, treatments, and covariates, why do they need to be jointly iid?
I read in a footnote in a paper that for the observational study setting in causal inference to hold, the outcomes, treatments and covariates must be iid. Specifically, if $Y_i$, $T_i$, $X_i$ are ...
0
votes
1answer
52 views
Follow up medical study with missing data
I am analyzing some patient data for a medical study that has a duration of several years.
Once a year, the patients are expected to visit the doctor, where they get four treatments, say A, B, C, D. ...
1
vote
0answers
19 views
What process should be used to test an ON-OFF-ON-OFF experiment?
The problem:
I have N (~500) black boxes which each receive an input and output a noisy reward signal. The reward signal is non-stationary and heteroscedastic, but can be assumed stationary over short ...
0
votes
1answer
36 views
Understanding Residual plots and how to transform my data to not violate the normal assumption
Hello I would like to know, whether my residual vs fitted plot is okay ie does not violate the normality assumption
The design is a random block design
as well as the histogram seems to be binomial ...
0
votes
1answer
27 views
Partially crossed design and how to apply the mixed model?
I have a list of patient samples that were tested at 3 clinical sites; for site 1, reagent lots 1 and 2 were applied; for site 2, reagent lots 1 and 3 were applied; for site 3, reagent lots 2 and 3 ...
0
votes
0answers
24 views
Non-orthogonal experimental design and model selection
I am working on designing some chemical experiments, with the goal (for now) to optimize reaction yield.
I intend to use principal component scores in order to investigate solvents, Lewis acids etc. ...
0
votes
0answers
7 views
How to decide sample size from Live/new data to evaluate a trained model performance
I have a dataset, partitioned that into training/holdout/test set using 80/10/10 rule. I trained/tested a xgboost binary classifier and deployed that into production. What sample size of live/new data ...
2
votes
0answers
23 views
Why do we discuss ANOVA and completely randomized design together?
It is common that textbook discusses completely randomized design and ANOVA together. However, it is not clear to me at all how does CRD "translate" to ANOVA. For example, does CRD imply that the ...
0
votes
0answers
31 views
WIth Latin square design, is this situation valid?
I have no background in statistics so I would like to know whether I get it right:
Q: suppose I evaluate 3 Teaching methods in 3 schools and in 3 classes (First, Second and Third).
I know that the ...
0
votes
0answers
7 views
Implications of bootstrap resampling on optimal DOE evaluation
Background:
I have a set of experiments that include some D-optimal DOE points, and I am trying to use bootstrap-resampling to get an estimate of the value of having those points in the analysis ...
0
votes
0answers
10 views
significance test by confidence interval in complete randomization design(CRD)
In page 91 of this book, Design and analysis of experiments by Montgomery , the author stated after Equation (3.30) that:
Clearly, if the confidence interval in Equation 3.30
includes zero, ...
0
votes
0answers
22 views
Principal component analysis with grouped data and few replicates
I am relatively new to principal component analysis so I am hoping someone can help me a bit.
I have a dataset in which I have measured 6 yield components (total yield, grain yield, no spikes, no. ...
0
votes
1answer
33 views
How can I test whether an individual treatment mean significant or not?
Suppose by complete randomization design, I reach into a decision that there is at least difference between two treatment means. That is, my hypothesis is
$$H_o:\mu_1=\mu_2=\mu_3=\mu_4=\mu_5$$
$$...
0
votes
0answers
12 views
Mixed cross-over & between-groups (parallel) design?
I would like to design an experiment that tests the efficacy of 3 different interventions upon 2 different populations of subjects (patients and controls).
I initially thought of a cross-over design,...
1
vote
1answer
40 views
Nested mixed model with longitudinal data and variables with very few observations
I am doing my first data analysis and I have a hard time translating the experiment design to the model I want to fit. I have a couple of basic questions about the overall coding of the model, and a ...
0
votes
0answers
25 views
What model can I go with when I have factors that cannot be changed?
I've been tasked with analyzing how several factors effect the impact resistance of a plastic molded part... unfortunately I have no experience doing much statistical work beyond typical linear fits. ...
0
votes
0answers
3 views
Quasi research design in 3 set ups with different sample sizes
We are working on our research entitled the effect of class size on the mathematical performance of the students, there we subjected 3 classes with different class size and we try to make everything ...
0
votes
0answers
18 views
Is two-factor factorial design most appropriate in this situation?
Suppose there are two factors, A and B. There are 7 levels for factor A and 4 levels of ...
